|Publication number||US4906904 A|
|Application number||US 07/372,060|
|Publication date||Mar 6, 1990|
|Filing date||Jun 27, 1989|
|Priority date||Jun 27, 1989|
|Publication number||07372060, 372060, US 4906904 A, US 4906904A, US-A-4906904, US4906904 A, US4906904A|
|Inventors||Keith I. Tognoni|
|Original Assignee||Digital Equipment Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (9), Referenced by (10), Classifications (8), Legal Events (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention is directed to cathode-ray-tube flyback circuits and in particular to those of the type required by high-resolution displays.
In a cathode-ray tube (a "CRT"), the position of the electron beam on the screen is determined by the current that flows through deflection yokes. FIG. 1 is a simplified schematic diagram of the type of circuit typically employed to control yoke current. The yoke Ly is shown connected to a fixed source voltage V+ at one end and through a switch S to ground at the other. A diode D and capacitor C are connected in parallel with the switch S. The drawing includes a conceptual diode Ds to indicate that the switch S is a unidirectional electronic switch. A gating signal Vg controls the state of the switch S.
FIGS. 2a-d illustrate the operation of this conventional circuit. The operation is cyclical, and FIGS. 2a-d enter the cycle at a time t0 at which the gating signal assumes an actuation value Von. FIG. 2a represents the value of the gating signal.
FIG. 2b depicts the yoke current. For reasons that will become apparent as the description proceeds, the current flow through the yoke is negative at time t0 ; i.e., the current flows toward the positive source terminal. Diode D therefore conducts, clamping one end of the yoke Ly to ground. Because a fixed voltage V+ is applied across the yoke Ly at this time, the yoke current increases linearly, and the horizontal position of the beam accordingly moves across the CRT screen to produce the forward, visible trace. As FIG. 2b shows, the yoke current becomes positive at time t1, so the yoke current stops flowing through the diode D and starts flowing through the switch S at that time.
At time t2, the gating signal Vg assumes a ground-level, disabling value, and the switch S responds after a delay by beginning to open at time t3. As is illustrated by FIG. 2c, which depicts the switch current, the switch S is completely open after a delay td. This begins the horizontal retrace.
Specifically, the yoke current begins to flow through the capacitor C when the switch S opens. The voltage on capacitor C accordingly rises. FIG. 2d depicts the capacitor voltage. The capacitance of capacitor C is low enough that the voltage rise is rapid; the purpose of capacitor C is to reverse the yoke current during the horizontal retrace time, and it is preferable to keep the retrace time low.
The resulting waveform of the current through the yoke Ly between times t3 and t4 is one-half cycle of a short-period sinusoid. When the yoke current becomes negative, the voltage on capacitor C decreases, reaching zero at time t4, as FIG. 2d illustrates.
After the capacitor voltage reaches zero at time t4, the diode D becomes forward biased, clamping the lower end of the yoke to ground and thereby causing a fixed voltage across the yoke. The yoke current therefore increases linearly from a negative value. During this linear increase, the gating signal Vg returns to the activation value Von that it assumed at time t0, and the cycle repeats.
The foregoing discussion is equally applicable both to conventional CRTs and to high-resolution CRTs. For high-resolution CRTs, however, the design constraints on circuit components are more stringent.
Increased CRT resolution means an increased number of horizontal scans for each vertical scan. Moreover, to avoid flicker, the vertical scan rate in higher-resolution displays is typically higher than that in conventional television. It is at least 60 Hz and preferably 66 Hz or even 80 Hz. The horizontal scan frequency is therefore much higher than the 15.75 kHz American television standard; it is typically at least 64 kHz and can exceed 100 kHz. The retrace time (tfly) must therefore be shortened from the 11 microseconds common in conventional television to a value in the range of 1.5-3.0 microseconds. This means that the peak capacitor voltage--and thus the voltage that the electronic switch can withstand--must increase. Additionally, field-effect transistors ("FETs") are the transistors of choice for such applications, and, for a given chip size, the on resistances of FETs increase faster than the squares of their breakdown voltages. The resultant resistance can detract from the linearity of the yoke-current curve.
One solution to this problem has been to configure the switch as a plurality of transistors connected in series; the total voltage that such a switch can withstand is proportional to the number of transistors, and the total switch resistance increases only linearly with the required voltage. Although this approach has been followed with some success, a certain degree of switch complexity has heretofore resulted from the requirement that the different switch transistors, which are at widely different potentials in the off state, be controlled by a common gate signal.
FIG. 3 depicts a typical prior-art circuit that illustrates this complexity. A series combination of capacitors C1 and C2 embodies the current-reversing capacitor C of FIG. 1, while the switch corresponding to switch S of FIG. 1 includes transistors Q1 and Q2 as well as the remaining circuitry in FIG. 3, which enables transistors Q1 and Q2 to respond to a common gating signal.
As will be explained presently, a high signal at terminal 12 causes transistors Q1 and Q2 to conduct. The conduction of transistor Q2 completes a current path to ground from a supply node 14 through a resistor R1, a diode D1, and the parallel combination of a zener diode D2 and a capacitor C3. As a result, the capacitor C3 is charged to the zener voltage and can be thought of as acting as a voltage supply that imposes a voltage between "supply rails" 16 and 18.
The high signal at the common gating terminal 12 causes current to flow through resistor R2, diode D3, and resistor R3 and forward biases a further diode D4 so as to bias a pnp transistor Q3 to its non-conducting state. The result is a low-voltage output of a voltage divider R5 and R6, and this low voltage prevents transistor Q4 from conducting. The resultant high voltage at the collector of transistor Q4 keeps transistor Q5 turned off and thereby isolates the gate of transistor Q1 from the lower "rail" 18 while permitting conduction through transistor Q6. Transistor Q6 thus provides a path from the high "rail" 16 through resistor R4 and transistor Q6 to the gate of transistor Q1, which accordingly conducts.
Similar circuitry causes transistor Q2 to conduct. There is thus very little potential difference between ground and the drain of transistor Q1, so the voltage across the series combination of capacitors C1 and C2 is nearly zero.
To turn transistors Q1 and Q2 off and thus open the switch, the signal at the common gating terminal 12 assumes a low, disablement value and thus allows the voltage across resistor R3 to drop low enough to turn on transistor Q3. The resultant voltage output from voltage divider R5 and R6 turns on transistor Q4, whose resultant low collector voltage reverses the states of capacitors Q5 and Q6 so that the base of Q1 now becomes isolated from the upper "rail" 16 and has its gate-to-source charge drawn off by transistor Q5. Transistor Q1 thus stops conducting.
By a similar mechanism, transistor Q2 also turns off, so capacitors C1 and C2 are allowed to charge rapidly. The junction of capacitors C1 and C2 is connected to that of transistors Q1 and Q2 so that the switch voltage is divided evenly between the transistors if the capacitances of capacitors C1 and C2 are equal. Consequently, no individual transistor is required to withstand the total switch voltage.
As capacitor C2 charges, the voltage at the lower "rail" 18 rises rapidly above those at the gating and supply terminals 12 and 14. However, diodes D1 and D3 isolate the power and gating terminals 14 and 12 from this high voltage, while capacitor C3 keeps the upper "rail" 16 at the level required to provide the necessary bias voltages for the various devices in the illustrated gate circuitry. The transistors thus remain non-conducting until the gating signal again assumes its activation value.
A couple of observations are in order concerning the circuitry of FIG. 3. The first is that the switch transistor circuit within dashed lines 19 is a four-terminal device; in addition to the control and power terminals--i.e., the base of transistor Q3 and the gate and drain of transistor Q1--circuit 19 includes a supply terminal ("rail" 16) to receive power to provide bias voltages. Although this terminal is effectively isolated during the off state of transistor Q1, it must be connected during the on state to recharge capacitor C3.
The second observation is that the relative complexity of the gate circuit not only has the potential for introducing delay but can also introduce differences between the turn-off times of the switch transistors. This can extend the switch turn-off time td (FIG. 2c) and thus extend the retrace time and increase timing uncertainty.
The present invention is a flyback circuit in which the transistor circuits that form the switch are simple three-terminal devices. Specifically, each transistor circuit, or module, includes two power terminals and a control terminal but no fourth terminal by which to apply bias voltage to the circuitry within the module.
In the case of FET embodiments, for instance, the power terminals are the transistor source and drain terminals, and the control terminal is connected to the transistor gate terminal. In addition to the transistor, the module needs to include only a turn-off circuit connected between the gate and the source to aid in turning the transistor off rapidly when the gating signal assumes is disabling value.
All the modules are connected to a common gating terminal, and this connection is provided, in the case of at least all of the modules but one, through a diode oriented to conduct in the direction required to convey an activation signal from the common gating terminal. The diode prevents current flow in the other direction, however, so it isolates the common gating terminal from the high switch voltages that can prevail when the switch is open.
These and further features and advantages of the present invention are described in connection with the accompanying drawings, in which:
FIG. 1 is a schematic diagram of a simplified version of a conventional flyback circuit;
FIGS. 2a-d depict waveforms at various points in the circuit of FIG. 1;
FIG. 3 is a schematic diagram of an exemplary prior-art switch for a flyback circuit;
FIG. 4 is a schematic diagram of one embodiment of the present invention; and
FIG. 5 is a schematic diagram of an alternate embodiment of the present invention.
FIG. 4 depicts a deflection circuit in which the capacitor C of FIG. 1 is embodied in a series combination of three capacitors C4, C5, and C6, and the switch S is embodied in a series combination of three FETs Q7, Q8, and Q9 and associated circuitry. The switch includes three identical transistor-circuit modules 20, 22, and 24. Module 22 is connected to a common gating terminal 26 by means of a diode D6, while module 22 is similarly connected by means of another diode D7. Module 24 is connected directly to the common gating terminal 26 without the interposition of a diode. Although the drawings depict an arrangement that includes three modules, the concept of the present invention is equally applicable to two, four, or more modules.
Each of the modules 20, 22, and 24 is a simple three-terminal device. In the case of module 20, for instance, the two power terminals are the gate and drain terminals of Q7, while the control terminal is on the node 28 to which diode D6 is connected. Unlike the circuit of FIG. 3, that of FIG. 4 requires no fourth terminal to provide bias voltages for the gate circuitry of transistors Q7, Q8, and Q9. Moreover, each module is very simple in design so that delay and variations in turn-on time are easily minimized.
In operation, an activation voltage of, say, 10 volts is applied to the common gating terminal 26 to keep the transistors turned on and the switch closed. Specifically, the activation voltage at the gate of transistor Q9 causes it to conduct, so the source terminal of transistor Q8 is approximately at ground potential. The activation signal at terminal 26 accordingly forward biases diode D7, causing current to flow through a resistor R8 and apply a turn-on voltage to the gate of transistor Q8. Therefore, the voltage at the source of transistor Q7 is also near ground, so transistor Q7 also conducts, and the switch is thus open.
To turn the transistors off and thus open the switch, the voltage at terminal 26 is returned to a low level, approximately ground. The charge on the gate-to-source capacitance of transistor Q7 initially keeps it turned on, but a fast-turn-off resistor R7 provides a path for rapidly dissipating the charge on this capacitance so that transistor Q7 quickly stops conducting. By a similar mechanism, transistors Q8 and Q9 also turn off, and capacitors C4, C5, and C6 therefore charge in the manner described in connection with FIG. 1.
The potential at the source terminal of transistor Q7 thus rises to a high level, as does that of the control terminal 28 of module 20, but the reverse bias of diode D6 isolates the common gating node of 26 from this high voltage. Diode D7 performs a similar function in isolating common gating terminal 26 from the high voltages in module 22. A similar diode is not necessary for module 24, because the gate voltage of transistor 29 does not rise as capacitors C4, C5, and C6 charge.
As was discussed in the background in connection with FIGS. 2a-d, the switch is to be closed again after the capacitors C4, C5, and C6 have been discharged. Accordingly, the source terminals of transistors Q7, Q8, and Q9 are nearly at ground potential at this time, so the activation potential at their gates causes them to conduct. The switch is thus closed again.
An activation voltage at terminal 26 would cause the switch to close even if a high voltage were present across the switch. In such a situation, the activation voltage, although not initially high enough to turn on transistors Q7 and Q8, would turn on transistor Q9 so that capacitor C6 would discharge. The voltage at the source terminal of transistor Q8 would therefore drop to a value low enough to permit the activation potential to turn on transistor Q8. By a similar mechanism, transistor Q8 would cause transistor Q9 to conduct.
FIG. 5 is an alternate embodiment of the present invention. The arrangement of FIG. 5 includes components L1', D5', C4', C5', and C6', whose functions are the same as those of the corresponding unprimed components in the arrangement of FIG. 4. However, the connections to yoke L1' in the circuit of FIG. 5 are similar to those in most conventional circuits and differ from those to yoke L1 in the simplified circuit of FIG. 4. The purpose of the difference is to modify the shape of the current curve during the forward trace. Rather than the strictly liner curve illustrated in FIG. 2b, the yoke current of the circuit of FIG. 5 follows a curve that tapers somewhat near the ends so as to overcome display distortions that would otherwise be introduced by the geometry of the CRT.
For this purpose, the fixed V+ voltage source of FIG. 4 is replaced with a large capacitor C7 whose average voltage is maintained by a fixed V+ voltage source acting through a choke coil L2. Capacitor C7 performs a function similar to that of the fixed source of FIG. 4, but its voltage increases in response to the negative yoke current during the first half of the horizontal trace and decreases in response to the positive yoke current during the second half so that the rate of charge of the yoke current is less at the ends of the trace than it is in the middle of the trace.
The inductance of the choke coil L2 is high enough that it can be considered an a.c. open circuit. Aside from its function of maintaining an average voltage on capacitor C7, therefore, the voltage source V+ can be considered isolated from the circuit. The operation of yoke circuit of FIG. 5 is thus the same as that of the FIG. 1 circuit except for the shape of the yoke-current curve during the forward trace. Accordingly, the operation of this part of the circuit of FIG. 5 will not be separately described.
Like the embodiment of FIG. 4, that of FIG. 5 includes three transistor modules 20', 22', and 24'. Like the corresponding modules of FIG. 4, these modules receive their gating signal from a common gating terminal 26', although, for reasons of interchangeability, the arrangement of FIG. 5 includes an additional diode D8 to provide the connection between module 24' and the common gating terminal 26. As will become apparent as the description proceeds, the difference between the modules 20', 22', and 24' of FIG. 4 is that the fast-turn-off resistors R7, R8, and R9 of FIG. 4 are replaced with fast-turn-off circuits that present less load to the gating-signal source.
With an activation voltage at the common gating terminal 26', transistors Q7', Q8', and Q9' conduct, so the source terminal of Q7' is nearly at ground potential. Diode D6' is thus forward-biased to permit current flow through a bias resistor R10. This results in a voltage at the base of a pnp transistor Q10 that forward biases a bias diode D9 so as to prevent conduction through pnp transistor Q10. The cathode voltage of diode D9 is thus two diode drops below the activation potential at terminal 26', and this voltage appears at the gate of transistor Q7' and keeps it conducting. Transistors Q8' and Q9' similarly conduct, so the voltage across the capacitor circuit comprising capacitors C4', C5', and C6' is held to zero.
To open the switch, the signal at the common gating terminal 26' is lowered approximately to ground, and diode D6 is accordingly back biased. Because of the gate-to-source capacitance of transistor Q7', its gate initially remains at the voltage required to keep it conducting. However, because the activation voltage no longer drives the upper side of resistor R10 to a high level, transistor Q10 can conduct and thus provide a current path through a current-limiting resistor R11 to ground. The source-to-drain capacitance Q7' thus discharges rapidly, and transistor Q7' quickly stops conducting. Transistors Q8' and Q9' are turned off in a similar manner, so the switch opens.
From the foregoing description, it can be appreciated that transistor Q10 and bias resistor R10, together with the resistor R11 and diode D9 in the activation-signal path from diode D6' to the gate of transistor Q7', act as a fast-turn-off circuit in place of the fast-turn-off resistor R7 of FIG. 4. The reason for using transistor Q10 in place of resistor R7 of FIG. 4 is to reduce the current drawn from the gating source when the switch is held closed. The resistance of resistor R7 of FIG. 4 should be small so as to enable Q7 to discharge rapidly. However, this low resistance also draws substantial current from the gating-signal source when the switch is closed. In the arrangement of FIG. 5, resistor R10 does conduct when the switch is closed, but its resistance can be much higher than that of resistor R7; when the switch opens, resistor R10 needs to conduct only a small fraction of the current required to discharge the Q7 source-to-drain capacitance rapidly.
The purpose of the current-limiting resistor R11 is to set the discharge current that transistor Q10 draws. Clearly, transistor Q10 would draw less current in the absence of resistor R11, but it is desirable to have transistors Q7', Q8', and Q9' discharge at the same rate and thus turn off simultaneously, and it is easier to match the resistances of R11 and corresponding resistors in the other modules 22' and 24' than to match the characteristics of transistor Q10 with those of its corresponding turn-off transistors in the other modules.
In order to protect transistor Q7' from any possible damaging reverse voltage, a diode D10 is connected between the gate and the source of transistor Q7' to prevent such reverse voltages between the gate and the drain.
By employing the teachings of the present invention, it is possible to perform flyback-circuit switching through the use of a switch whose individual transistor-circuit modules are controlled by a common gating signal and require no separate bias-voltage supply terminal. It is thus apparent that the present invention constitutes a significant advance in the art.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4242714 *||Mar 21, 1979||Dec 30, 1980||Sony Corporation||Current control circuit|
|US4487457 *||Sep 28, 1982||Dec 11, 1984||Eaton Corporation||Gating circuit for combining series and parallel connected FETs|
|US4500801 *||Jun 21, 1982||Feb 19, 1985||Eaton Corporation||Self-powered nonregenerative fast gate turn-off FET|
|US4535273 *||Aug 17, 1984||Aug 13, 1985||Zenith Electronics Corporation||Transformerless switching circuit for driving a horizontal output transistor|
|US4556825 *||Sep 9, 1983||Dec 3, 1985||Intergraph Corporation||Switching circuit|
|US4580063 *||Dec 1, 1982||Apr 1, 1986||Sgs-Ates Componenti Elettronics S.P.A.||Circuit for the protection of IGFETs from overvoltage|
|US4692693 *||Jan 24, 1985||Sep 8, 1987||Commissariat A L'energie Atomique||Apparatus for measuring low currents with high dynamics|
|US4751408 *||Aug 26, 1986||Jun 14, 1988||Thomson-Csf||Voltage-switching device|
|US4777387 *||Feb 21, 1984||Oct 11, 1988||International Rectifier Corporation||Fast turn-off circuit for photovoltaic driven MOSFET|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US5177411 *||Jul 5, 1991||Jan 5, 1993||Sony Corporation||CRT correcting circuit|
|US5485062 *||Jan 14, 1994||Jan 16, 1996||Sony Corporation||Horizontal deflection circuit for a television set|
|US6573752||Jan 10, 2002||Jun 3, 2003||Dialog Semiconductor Gmbh||High voltage push-pull driver on standard CMOS|
|US8994442 *||Jun 18, 2013||Mar 31, 2015||Fairchild Semiconductor Corporation||Switching circuit and controller circuit|
|US9252651 *||Jun 20, 2012||Feb 2, 2016||Synaptics Incorporated||High voltage driver using medium voltage devices|
|US20130002149 *||Jun 20, 2012||Jan 3, 2013||Synaptics Incorporated||High voltage driver using medium voltage devices|
|US20130342262 *||Jun 18, 2013||Dec 26, 2013||Fairchild Semiconductor Corporation||Switching circuit and controller circuit|
|US20150249448 *||Feb 28, 2014||Sep 3, 2015||Infineon Technologies Austria Ag||Electronic Circuit Operable as an Electronic Switch|
|US20160155737 *||Aug 28, 2015||Jun 2, 2016||Globalfoundries Singapore Pte. Ltd.||Rc-stacked mosfet circuit for high voltage (hv) electrostatic discharge (esd) protection|
|EP1326337A1 *||Jan 3, 2002||Jul 9, 2003||Dialog Semiconductor GmbH||High voltage push-pull driver on standard CMOS|
|U.S. Classification||315/408, 315/397|
|International Classification||H03K17/10, H03K4/62|
|Cooperative Classification||H03K17/102, H03K4/62|
|European Classification||H03K4/62, H03K17/10B|
|Jun 27, 1989||AS||Assignment|
Owner name: DIGITAL EQUIPMENT CORPORATION, 111 POWDERMILL ROAD
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNOR:TOGNONI, KEITH I.;REEL/FRAME:005097/0884
Effective date: 19890623
|Sep 2, 1993||FPAY||Fee payment|
Year of fee payment: 4
|Sep 5, 1997||FPAY||Fee payment|
Year of fee payment: 8
|Aug 16, 2001||FPAY||Fee payment|
Year of fee payment: 12
|Jan 9, 2002||AS||Assignment|
Owner name: COMPAQ INFORMATION TECHNOLOGIES GROUP, L.P., TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DIGITAL EQUIPMENT CORPORATION;COMPAQ COMPUTER CORPORATION;REEL/FRAME:012447/0903;SIGNING DATES FROM 19991209 TO 20010620
|Jan 21, 2004||AS||Assignment|
Owner name: HEWLETT-PACKARD DEVELOPMENT COMPANY, L.P., TEXAS
Free format text: CHANGE OF NAME;ASSIGNOR:COMPAQ INFORMATION TECHNOLOGIES GROUP, LP;REEL/FRAME:015000/0305
Effective date: 20021001